Professional Documents
Culture Documents
03 20 18 Party Identification CORRECTED PDF
03 20 18 Party Identification CORRECTED PDF
202.419.4372
www.pewresearch.org
RECOMMENDED CITATION
Pew Research Center, March, 2018, “Wide
Gender Gap, Growing Educational Divide in Voters’
Party Identification”
PEW RESEARCH CENTER
www.pewresearch.org
PEW RESEARCH CENTER
Since 2014, the last midterm election year, there have been
notable changes in party identification among several groups of
voters. And as we noted in our 2016 report on party affiliation,
the composition of the Republican and Democratic electorates
are less alike than at any point in the past quarter-century. Note: Based on registered voters.
Source: Annual totals of Pew Research
Center survey data (U.S. adults).
PEW RESEARCH CENTER
www.pewresearch.org
2
PEW RESEARCH CENTER
Persistent gender gap. For decades, women have been more likely than men to identify as
Democrats or lean Democratic. But today, a 56% majority of women identify as Democrats or lean
Democratic, while 37% affiliate with or lean toward the GOP. The share of women identifying as
Democrats or leaning Democratic is up 4 percentage points since 2015 and is at one of its highest
points since 1994. Among men, there has been less recent change: 48% identify with the
Republican Party or lean Republican, while 44% are Democrats or lean Democratic. That is
comparable to the balance of leaned party identification since 2014.
Record share of college graduates align with Democrats. Voters who have completed college make
up a third of all registered voters. And a majority of all voters with at least a four-year college
degree (58%) now identify as Democrats or lean Democratic, the highest share dating back to
1994. Just 36% affiliate with the Republican Party or lean toward the GOP. The much larger group
of voters who do not have a four-year degree is more evenly divided in partisan affiliation. And
voters with no college experience have been moving toward the GOP: 47% identify with or lean
toward the Republican Party, up from 42% in 2014.
Continued racial divisions in partisan identification. About half of white voters (51%) identify with
the GOP or lean Republican, while 43% identify as Democrats or lean Democratic. These figures
are little changed from recent years. By contrast, African American voters continue to affiliate with
the Democratic Party or lean Democratic by an overwhelming margin (84% Democrat to 8%
Republican). Hispanic voters align with the Democrats by greater than two-to-one (63% to 28%),
while Asian American voters also largely identify as Democrats or lean Democratic (65%
Democrat, 27% Republican).
Larger differences among whites by education. Most white voters with at least a four-year college
degree (53%) affiliate with the Democratic Party or lean Democratic; 42% identify as Republicans
or lean Republican. As recently as two years ago, leaned partisan identification among white
college graduates was split (47% Democrat, 47% Republican). Majorities of white voters with some
college experience but who do not have a degree (55%) and those with no college experience (58%)
continue to identify as Republicans or lean Republican.
Millennials, especially Millennial women, tilt more Democratic. As noted in our recent report on
generations and politics, Millennial voters are more likely than older generations to affiliate with
the Democratic Party or lean Democratic. Nearly six-in-ten Millennials (59%) affiliate with the
Democratic Party or lean Democratic, compared with about half of Gen Xers and Boomers (48%
each) and 43% of voters in the Silent Generation. A growing majority of Millennial women (70%)
affiliate with the Democratic Party or lean Democratic; four years ago, 56% of Millennial women
did so. About half of Millennial men (49%) align with the Democratic Party, little changed in
www.pewresearch.org
3
PEW RESEARCH CENTER
recent years. The gender gap in leaned party identification among Millennials is wider than among
older generations.
The nation’s changing demographics – and shifting patterns of partisan identification – have had a
profound impact on the makeup of the Democratic and Republican electorates.
Across several dimensions – race and ethnicity, education and religious affiliation – the profile of
Democratic and Democratic-leaning registered voters has changed a great deal over the past two
decades. The composition of
Republican and Republican- Democratic voters have become more racially diverse
leaning voters has shown less since late 1990s; less change among GOP voters
change. % of registered voters who are …
Race and ethnicity
While a majority of voters White Black Hispanic Asian Other
(69%) are white non- All Rep/ Dem/
voters Lean Rep Lean Dem
Hispanics, nonwhite voters
now make up an increasing 6 5
10 8 6 9
10 17 12
share of all voters: 29% of 11
11 18
19
registered voters are African
American, Hispanic or Asian
92 88
83 83
American or belong to 76 69 75
67
59
another race, up from 16% in
1997. Nonwhites constitute
nearly four-in-ten
'97 '07 '17 '97 '07 '17 '97 '07 '17
Democratic voters (39%),
Note: Based on registered voters.
compared with 24% two Source: Annual totals of Pew Research Center survey data (U.S. adults).
decades ago. The GOP PEW RESEARCH CENTER
coalition also has become
more racially and ethnically
diverse, but nonwhites make up only 14% of Republican voters, up from 8% in 1997.
The educational makeup of the two parties’ electorates also has changed substantially over the past
two decades. When race and education are taken into account, white voters who do to not have a
college degree make up a diminished share of Democratic registered voters. White voters who do
not have a four-year degree now constitute just a third of Democratic voters, down from 56% two
www.pewresearch.org
4
PEW RESEARCH CENTER
decades ago. By contrast, non-college white voters continue to make up a majority of Republican
and Republican-leaning registered voters (59% now, 66% in 1997).
The share of Democratic voters describing their political views as liberal has increased steadily
since 2000. Republicans’ ideological views have changed little over past decade, but the share of
Republicans identifying as conservatives rose between 2000 and 2008.
Currently, nearly half of Democratic and Democratic-leaning registered voters (46%) say they are
liberal, while 37% identify as moderates and 15% say they are conservatives. A decade ago, more
Democrats described their views as moderate (44%) than liberal (28%), while 23% said they were
conservative.
Conservatives have long constituted the majority among Republican and Republican-leaning
registered voters. Roughly two-thirds of Republicans (68%) characterize their views as
conservative, while 27% are moderates and 4% are liberals. While there has been little change in
Republicans’ self-described ideology in recent years, the share calling themselves conservatives
rose from 58% in 2000 to 65% eight years later.
68
58
44 46
38 38
38 32 37
33 28
27
26 23
17 15
6 4
| | |
www.pewresearch.org
5
PEW RESEARCH CENTER
There continue to be fundamental differences '94 '98 '02 '06 '10 '14 '17
in the partisan orientation of different
demographic groups, and in many cases these Note: Based on registered voters.
Source: Annual totals of Pew Research Center survey data (U.S.
gaps have grown wider in recent years. For adults).
instance, gender, generational, geographic and PEW RESEARCH CENTER
www.pewresearch.org
6
PEW RESEARCH CENTER
As has been the case for more than two decades of Pew Research Center surveys, women are
significantly more likely than
men to associate with the
Share of women who identify with or lean toward
Democratic Party. While the
Democratic Party has risen since 2015
gender gap has changed little
% of registered voters who identify as …
in recent years, it is as wide
Rep34
35
28 Dem 4228
26 Ind Rep/Ln
51Rep
39 Dem/Ln
51
43 Dem
as it has been at any point
Men
during this period: Among
registered voters, 56% of
women affiliate with or lean 52 48
42
35 39 44
toward the Democratic Party, 34 28
28 26
compared with 44% of men.
| |
From 2010 through 2015, '94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
about half of women (51%-
Women
52%) identified with or
leaned toward the 56
Democratic Party. But the 48
37 39
share of women who identify 31 32 42 37
27 25
with or lean to the
Democratic Party has risen in | |
recent years, to 54% in 2016 '94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
and 56% in 2017. The Note: Based on registered voters.
partisan breakdown of men is Source: Annual totals of Pew Research Center survey data (U.S. adults).
this period.
The Democratic gains among women have not come from increased affiliation with the party.
Overall, the proportion of women voters who identify with (rather than lean toward) the
Democratic Party has remained relatively constant for the past 25 years (in 1994, 37% of women
said they identified with the Democratic Party, compared with 39% in 2017).
www.pewresearch.org
7
PEW RESEARCH CENTER
There are sizable and long-standing racial and ethnic differences in partisan affiliation, and they
have shifted only modestly in recent years.
share of whites identifying as '94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
increase in Democratic- '94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
overwhelmingly Democratic: '94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
84% identify with or lean Notes: Based on registered voters. Whites and blacks include only those who are not
Hispanic; Hispanics are of any race. Data for Hispanics shown only for years in which
toward the Democratic Party. interviews were conducted in both English and Spanish.
Source: Annual totals of Pew Research Center survey data (U.S. adults).
Just 8% of black voters
PEW RESEARCH CENTER
identify in some way with the
Republican Party.
www.pewresearch.org
8
PEW RESEARCH CENTER
While black voters remain solidly Democratic, identification with the Democratic Party has
declined modestly in recent years: About two-thirds of African Americans have identified as
Democrats in the last several years, down slightly from the first half of Barack Obama’s
presidency, when about three-quarters affiliated with the Democratic Party.
By more than two-to-one (63% to 28%), Hispanic voters are more likely to affiliate with or lean
toward the Democratic Party than the GOP. The overall balance of partisan orientation among
Hispanics is little changed over the last decade.
There is a similar balance of partisanship among Asian American registered voters: 65% identify
with the Democratic Party or lean Democratic, compared with 27% who identify as or lean
Republican.
'94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
and 33% identified with or
leaned toward the Republican Notes: Based on registered voters. Asians include only those who are not Hispanic and are
English speaking.
Party. (Note: Only English- Source: Annual totals of Pew Research Center survey data (U.S. adults).
speaking Asian American PEW RESEARCH CENTER
The share of Asian American voters who identify as Republican is now only 12%. While this is little
changed in the last few years, it represents the continuation of a longer trend in declining
Republican affiliation among Asian voters. Among Asians, identification with the Democratic
Party has remained relatively stable over this period. The share of Asian voters who say they are
political independents has risen steadily since 1998, reflecting a more general trend among all
voters.
www.pewresearch.org
9
PEW RESEARCH CENTER
Men Women
For instance, there is a 9- Among whites Among blacks
Women
percentage-point gender gap 87
83 79
among white voters: While 79
Men
48% of white women affiliate Women
48
44 37
with or lean toward the 35 Men
Democratic Party, 37% of
white men do so. Similarly,
| |
'94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
there is an 8-point gender
Among Hispanics
gap among black voters (87%
of black women vs. 79% of Women
black men), as well as among 61
66
58
Hispanic voters (66% of 55 Men
women vs. 58% of men).
Notes: Based on registered voters. Whites and blacks include only those who are not
Hispanic; Hispanics are of any race. Data for Hispanics shown only for years in which
interviews were conducted in both English and Spanish. Asian Americans by gender not
shown because of insufficient sample sizes.
Source: Annual totals of Pew Research Center survey data (U.S. adults).
PEW RESEARCH CENTER
www.pewresearch.org
10
PEW RESEARCH CENTER
Higher educational attainment is increasingly associated with Democratic Party affiliation and
leaning. At the same time, those without college experience – once a group that tilted more
Democratic than Republican
– are roughly divided in their
partisan orientation. College graduates continue to shift toward the
Democratic Party
% of registered voters who identify as …
These twin shifts have
resulted in the widest Rep34
35
28 Dem 42
28
26 Ind Rep/Ln
51Rep
39 Dem/Ln
51
43 Dem
HS or less
educational gap in partisan
identification and leaning
seen at any point in more 47 47
37 33 42 45
than two decades of Pew 30 32
29 30
Research Center surveys.
| |
'94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
In 1994, 39% of those with a
Some college
four-year college degree (no
postgraduate experience)
50 47
identified with or leaned 35 39
40 45
toward the Democratic Party 31 29
30 28
and 54% associated with the
Republican Party. In 2017, | |
'94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
those figures were exactly
College graduate (no postgraduate experience)
reversed.
54 54
Democratic gains have been 39 38
32 35 39 39
even more pronounced
27
24
among those who pursue
postgraduate education. In | |
'94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
1994, those with at least some
Postgraduate experience
postgraduate experience were
63
evenly split between the
47
Democratic and Republican 41 45
33 38
parties. Today, the 32
32 31
19
Democratic Party enjoys a
roughly two-to-one | |
'94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
advantage in leaned partisan
Note: Based on registered voters.
Source: Annual totals of Pew Research Center survey data (U.S. adults).
PEW RESEARCH CENTER
www.pewresearch.org
11
PEW RESEARCH CENTER
identification. While some of this shift took place a decade ago, postgraduate voters’ affiliation
with and leaning to the Democratic Party have grown substantially just over the past few years,
from 55% in 2015 to 63% in 2017.
By contrast, Republicans have been gaining ground over the past several years with those who do
not have bachelor’s degrees. Among those with no more than a high school education, 47% affiliate
with the GOP or lean Republican, while 45% identify as Democrats or lean Democratic. Democrats
held a significant advantage among voters with a high school degree or less education for much of
the late 1990s through early 2000s, and as recently as 2014 (47% Democratic, 42% Republican).
These overall patterns in education and partisanship are particularly pronounced among white
voters. While the GOP has held significant advantages over the Democratic Party among white
college graduates without
postgraduate experience over
Divide in partisanship continues to widen between
much of the past two
white voters with and without a college degree
decades, these voters are
% of white registered voters who identify as …
divided in their partisanship
today. Rep/Ln
51Rep
39 Dem/Ln
51
43 Dem
HS or less Some college
GOP. As recently as 2015, '94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
51% of white voters with a
Four-year college degree Postgrad experience
college degree aligned with
the Republican Party, 59 59
49 47
compared with 43% for the 46 46
Democratic Party. 37
34
postgraduate experience, the '94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
Democratic advantage has Notes: Based on white registered voters. Whites include only those who are not Hispanic.
grown. In 2017, 59% of white Source: Annual totals of Pew Research Center survey data (U.S. adults).
www.pewresearch.org
12
PEW RESEARCH CENTER
leaned toward the Republican Party; as recently as 2015 that balance was slightly narrower (52%
to 41%).
By contrast, white voters with no more than a high school education have moved more to the GOP
over the last 10 years, though there has been little change since 2015. As recently as 2009, these
voters were divided in leaned partisanship. Since then, Republicans have held significant
advantages, including a 23-percentage-point lead in 2017 (58% Republican, 35% Democratic).
www.pewresearch.org
13
PEW RESEARCH CENTER
Democratic tilt since they '94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
among Millennial voters '94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
Democratic, 37% tilted '94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
(44% vs. 39% of Gen Xers, '94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
www.pewresearch.org
14
PEW RESEARCH CENTER
Generation X voters (born 1965 to 1980) are more divided in their partisan attachments, but also
tilt toward the Democratic Party (48% identify as or lean Democratic, 43% identify as or lean
Republican). The balance of leaned partisan identification among Gen X voters has been relatively
consistent over the past several years. Baby Boomer voters (born 1946 to 1964) are nearly evenly
divided (48% identify as or lean Democratic, 46% Republican).
The Silent Generation (born 1928 to 1945) is the only generational group that has more GOP
leaners and identifying voters than Democratic-oriented voters. About half (52%) of Silent
Generation voters identify with or lean toward the Republican Party, a larger share than a decade
ago; 43% identify with or lean toward the Democratic Party.
www.pewresearch.org
15
PEW RESEARCH CENTER
Millennial men. '94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
Democratic advantage among '94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
Millennial women was a Boomer men Boomer women
narrower – but still
substantial – 21 percentage
53 51 53
46
points, compared with 47 43 43 41
38
points today. The balance of
partisanship among
| |
Millennial men was similar in '94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
2014 as it is today (50% Silent men Silent women
Democratic vs. 40%
Republican). 57
49 50 48
41 46
38 41
| |
'94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
Note: Based on registered voters.
Source: Annual totals of Pew Research Center survey data (U.S. adults).
PEW RESEARCH CENTER
www.pewresearch.org
16
PEW RESEARCH CENTER
overwhelmingly Democratic
| |
'94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
in their leanings, while whites White Gen X Nonwhite Gen X
are more divided. Among
66 64
white voters, Millennials are 57
52
the only generation in which 41
37
the share of Democrats and 27
24
Democratic leaners (52%) is | |
greater than the share of '94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
Republicans and Republican White Boomer Nonwhite Boomer
leaners (41%). 71
68
53 53
41
Among older generations of 37
whites – but particularly 21 22
'94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
voters align with the GOP
White Silent Nonwhite Silent
than the Democratic Party.
80 75
59
White Silent Generation 49
41 36
voters have moved toward
19
the GOP in recent years. 12
Today, 59% identify with or
| |
'94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
lean to the GOP, up from
Notes: Based on registered voters.
43% a decade ago. Whites include only those who are not Hispanic; nonwhites include Hispanics.
Sources: Annual totals of Pew Research Center survey data (U.S. adults).
PEW RESEARCH CENTER
www.pewresearch.org
17
PEW RESEARCH CENTER
White Gen Xer and Boomer voters have remained relatively stable in their partisan makeup in
recent years. In both generations, the GOP enjoys a similar modest advantage (11 percentage
points among white Gen Xers, 12 points among white Boomers); these are similar to the balances
in 2014.
www.pewresearch.org
18
PEW RESEARCH CENTER
White evangelical Protestants remain one of the most reliably Republican groups of voters, and the
GOP’s advantage among this
segment of the population Steady increase in GOP advantage in leaned partisan
has continued to grow in identification among white evangelical voters
recent years: 77% of white % of registered voters who identify as …
evangelical voters lean Rep34
35
28 Dem 4228
26 Ind Rep/Ln51Rep
39 Dem/Ln
51
43 Dem
toward or identify with the White non-Hispanic evangelical Protestant
Republican Party, while just
77
18% have a Democratic 61
56
orientation. 44
27 29 31
25 18
11
White mainline Protestant | |
voters are more divided in '94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
with the Democratic Party. '94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
the Democratic Party. '94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
Notes: Based on registered voters.
Whites and blacks include only those who are not Hispanic.
Source: Annual totals of Pew Research Center survey data (U.S. adults).
PEW RESEARCH CENTER
www.pewresearch.org
19
PEW RESEARCH CENTER
'94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
White Catholic voters now
White non-Hispanic Catholic
are more Republican (54%)
than Democratic (40%).
While the partisan balance 54
45
34 36 45
among white Catholic voters 31 35 40
31 26
is little changed in recent
years, this group was more
| |
'94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
evenly divided in their
Hispanic Catholic
partisan loyalties about a
decade ago.
69
64
56 50
Hispanic Catholics, who 31 27
23 26
represent a growing share of 18 13
the Catholic population in the
| |
'94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
U.S., are substantially more
Notes: Based on registered voters. Whites include only those who are not Hispanic;
Democratic in their Hispanics are of any race. Data for Hispanics shown only for years in which interviews were
conducted in both English and Spanish.
orientation (64% of Hispanic Source: Annual totals of Pew Research Center survey data (U.S. adults).
Catholic voters affiliate with PEW RESEARCH CENTER
or lean to the Democratic
Party, 27% to the GOP).
www.pewresearch.org
20
PEW RESEARCH CENTER
independent (41%) as they '94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
'94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
Note: Based on registered voters.
Source: Annual totals of Pew Research Center survey data (U.S. adults).
PEW RESEARCH CENTER
www.pewresearch.org
21
PEW RESEARCH CENTER
'94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
Today nearly seven-in-ten
Note: Based on registered voters.
(68%) do so. Source: Annual totals of Pew Research Center survey data (U.S. adults).
PEW RESEARCH CENTER
www.pewresearch.org
22
PEW RESEARCH CENTER
'94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
Overall, those who live in Suburban counties
suburban counties are about
evenly divided in their 47
47
partisan loyalties (47% 38 42 45
34 30
31 28
Democratic, 45% 31
Republican), little changed
| |
over the last two decades. '94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
Rural counties
Voters in rural areas have
moved in a more Republican 54
45
direction over the last several 32 36 44 38
32 34
26
years. From 1999 to 2009, 31
rural voters were about
| |
equally divided in their '94 '98 '02 '06 '10 '14 '17 '94 '98 '02 '06 '10 '14 '17
partisan leanings. Today, Notes: Based on registered voters. Urban, suburban and rural based on National Center for
there is a 16-percentage-point Health Statistics classification system. Suburban counties include both suburban counties
and small metros.
advantage for the GOP Source: Annual totals of Pew Research Center survey data (U.S. adults).
among rural voters. PEW RESEARCH CENTER
www.pewresearch.org
23
PEW RESEARCH CENTER
were roughly evenly divided in their political preferences for much of the last two decades, in
recent years the Democratic Party has enjoyed a double-digit partisan advantage: Today, 54% of
white urban voters are Democrats or lean Democratic, while 41% identify with the GOP or lean
Republican.
www.pewresearch.org
24
PEW RESEARCH CENTER
www.pewresearch.org
25
PEW RESEARCH CENTER
The overall growth in the racial and ethnic diversity of voters has changed the composition of both
the Republican and Democratic parties. Yet the pace of change has been more pronounced among
Democrats and Democratic leaners.
White registered voters make up a declining share of the Democratic Party. In 1997, 75% of
Democratic and Democratic-leaning registered voters were white; that has dropped to a smaller
majority today (59%). Nonwhite voters now make up about four-in-ten Democratic voters (39%),
up from 24% in 1997.
The electorate continues to grow older, impacting the age composition of Democratic and
Republican voters. In 1997, the median age of all registered voters was 45; today the median age
has risen to 50.
Nearly six-in-ten Republican and Republican-leaning voters (57%) are ages 50 and older,
compared with 42% who are under 50. Among Democratic voters, a larger share are younger than
50 (53%) than 50 and older (46%).
Twenty years ago, the age profiles of the two parties were much more similar. At that time,
comparable majorities of both parties’ voters were younger than 50 (61% of Republicans, 57% of
Democrats).
In 1997, the median age of Republican voters was 43, while the median age of Democratic voters
was 46. Today, the median age of Republican voters has increased nine years, to age 52, while the
median age of Democratic voters has edged up to 48.
www.pewresearch.org
26
PEW RESEARCH CENTER
Over the past 20 years, the American electorate has become better educated, with the share of
college graduates rising. In 1997, 45% of all registered voters had no college experience; today that
share has fallen to 33%.
College grads make up a growing share of Democratic
And while those with no voters; less change among Republican voters
college experience was the % of registered voters who are …
largest category of voters two
Education
decades ago, today the 45or less
29
25
HS Some33
39
28
college 33grad+
College
electorate is evenly divided – 97 07 17
All Rep/ Dem/
a third each are college voters Lean Rep Lean Dem
graduates, have some college
experience but no degree and 25 33 33 28 35 28 24 31 39
have no more than a high
school diploma. 29 27
28 32 35 28
33 30
31
Despite these shifts,
45 49 41
Republican and Republican- 39 33 40 35 37 30
leaning voters are no more
likely to be college graduates '97 '07 '17 '97 '07 '17 '97 '07 '17
than was the case two
Education by Race
decades ago. And college
graduates make up a smaller Non-coll white Coll white Non-coll nonwhite Coll nonwhite
www.pewresearch.org
27
PEW RESEARCH CENTER
college graduates; 32% had some college experience; and 40% had no more than a high school
education. And in 2007, college graduates made up a greater share of Republican voters than is
currently the case (35% of all GOP voters then, 28% now).
The educational makeup of Democratic voters has changed substantially over the past 20 years.
Today, about four-in-ten Democrats (39%) have at least a college degree, up from 24% in 1997.
And while voters with no more than a high school education constituted the largest share of
Democratic voters 20 years ago, today college graduates make up the largest share.
Whites without a college degree remain the largest share of all registered voters, but their numbers
have been on the decline due to growing diversity and rising levels of education in the population.
In 1997, a majority of all registered voters (61%) were whites without a college degree. Over the
past 20 years, that share has fallen to 44%.
The share of whites with at least a college degree has edged up from 22% of registered voters in
1997 to 25% today. Among nonwhites, the share with a college degree or more education has more
than doubled, from just 3% in 1997 to 8% in 2017. And nonwhites without a college degree make
up a much larger proportion of the electorate today (21%) than 20 years ago (13%).
Combining race and education, Democratic voters are very different today than they were 20 years
ago. Today, non-college whites make up a third of all Democratic voters; they constituted a
majority of Democrats (56%) in 1997. Since then, the share of white Democrats with at least a
four-year degree has increased from 19% to 26%, and the share of nonwhite Democratic college
graduates has more than doubled (from 5% to 12%).
Whites who do not have a four-year college degree continue to make up a majority of Republican
voters, though a smaller majority than 20 years ago (59% now, 66% then). Whites with at least a
four-year degree constitute about a quarter of Republican voters (24%), little changed from 1997
(26%).
www.pewresearch.org
28
PEW RESEARCH CENTER
CORRECTION (Jan. 2, 2019): In the chart “Growing share of voters, especially Democrats, are
previously unaffiliated,” the percentages of registered voters who are “other Christian” and “all
others” have been updated to correct a data tabulation error.
Religiously unaffiliated voters now account for a third of Democratic voters, up from just 9% in
1997. In fact, they make up a larger proportion of Democrats than do white Protestants (33% vs.
18%). In 1997, 40% of Democratic voters identified as white Protestants (evangelical or mainline),
while just 9% were religiously unaffiliated. And white Catholics, who made up about one-in-five
Democrats then (22%), account for only 10% of Democrats now.
Republicans continue to be mostly made up of white Christians: A third of Republican voters are
white evangelical Protestants, which is little changed from 1997 (34%); 17% are white mainline
Protestants (28% in 1997); and 17% are white Catholics (20% then).
www.pewresearch.org
29
PEW RESEARCH CENTER
While religiously unaffiliated voters constitute a much smaller segment of Republican than
Democratic voters, the share of Republicans who do not identify with a religious denomination has
risen. Currently, 13% of Republicans do not identify with a religion, up from 5% two decades ago.
Acknowledgements
This report is a collaborative effort based on the input and analysis of the following individuals:
www.pewresearch.org
30
PEW RESEARCH CENTER
Methodology
The analysis of changes in party identification over time is based on a compilation of 257 surveys
and nearly 350,000 interviews among registered voters conducted by the Pew Research Center
from January 1994 to December 2017. These surveys are combined into one large data file that can
be sorted according to a range of demographic characteristics, with comparisons made across
different time periods. Yearly totals are calculated by combining all surveys for the calendar year,
with appropriate weights applied. The table below shows the number of surveys and interviews
conducted each year as well as the margin of error for each yearly sample.
Number of Margin
Year Surveys Sample Size of Error
1994 6 7,696 1.3
1995 4 5,079 1.6
1996 7 7,988 1.3
1997 9 8,907 1.2
1998 13 15,223 0.9
1999 10 11,102 1.1
2000 11 15,662 0.9
2001 10 9,985 1.1
2002 11 14,215 1.0
2003 12 11,658 1.1
2004 15 20,162 0.8
2005 13 17,142 0.9
2006 14 19,850 0.8
2007 12 16,881 0.9
2008 16 26,126 0.7
2009 13 18,469 0.8
2010 12 20,182 0.8
2011 12 15,644 0.9
2012 12 18,779 0.8
2013 10 12,849 1.0
2014 12 19,436 0.8
2015 8 13,332 1.0
2016 8 12,441 1.0
2017 7 10,245 1.1
www.pewresearch.org
31
PEW RESEARCH CENTER
and Orthodox Christians. The “All others” category includes Black Protestant
White Catholic
Jews, Muslims, Buddhists, Hindus, members of other faiths, and
Hispanic Catholic
those who declined to provide their religion; individually, each of
Religiously unaffiliated
these groups accounts for 2% or less of all registered voters.
Other Christian
Other Protestants
Other Catholics
Mormons
Orthodox Christians
All others
Jewish
Muslim
Buddhist
Hindu
Other non-Christian faiths
DK/Ref
PEW RESEARCH CENTER
www.pewresearch.org